Berlin 2024 – scientific programme

Parts | Days | Selection | Search | Updates | Downloads | Help

BP: Fachverband Biologische Physik

BP 34: Statistical Physics of Biological Systems III (joint session BP/DY)

BP 34.9: Talk

Friday, March 22, 2024, 12:00–12:15, H 2032

Two-stage adaptive evolution in a rugged yet highly-accessible fitness landscape model with delayed commitment — •Muhittin Mungan¹, Suman G. Das², and Joachim Krug¹ — ¹Institute for Biological Physics, University of Cologne, Köln, Germany — ²Institute of Ecology and Evolution, University of Bern, Bern, Switzerland

We study an empirically-motivated theoretical fitness landscape model of antibiotic-resistance evolution in bacteria [1]. The fitness of a genotype at any concentration depends on two parameters, the resistance level and the drugless growth rate, with a tradeoff between these. For intermediate concentrations fitness landscapes are rough while the fitness peaks are nevertheless highly accessible. Adaptation on such landscapes occurs in two stages. At first there is rapid accumulation of mutations and fast growth of fitness and resistance level along with a decrease in the drugless growth rate. Next there is slow growth of fitness and resistance level, and recovery of the drugless growth rate through the reversion of some of the mutations. We numerically demonstrate a robust pattern of evolution with qualitative features that are largely independent of specific model assumptions. The basins of individual fitness peaks overlap strongly, so that commitment to a peak is delayed substantially. An analytically tractable special case reproduces our findings rather well, shedding light on the nature of adaptive evolution, basins of fitness peaks and delayed commitment.

[1] S.G. Das et al. eLife 9:e55155 (2020)

Keywords: Fitness landscapes; Ruggedness and accessibility; Adaptive evolution; Disordered systems; Antibiotic resistance